The present invention relates generally to online management of historical data, and more particularly, to efficient reporting and analytics using online management of historical data.
As enterprise systems evolve and grow bigger, managing these systems is becoming increasingly difficult and complex. Manual management of enterprise systems can be costly, error-prone, time-consuming, and is not scalable. In recent years, much emphasis has been put on automating various system management tasks, such as discovery, provisioning, chargeback, configuration management, resource management, and end-to-end analysis. Many of these management tasks have to analyze historical data generated by the different components of the enterprise systems or applications running on the enterprise systems. Most realistic systems generate large amounts of historical data that may include information related to usage, performance, and/or configuration changes.
In order to manage these systems effectively, the management solutions for these systems should be able to scale and efficiently process large quantities of data. It is known that the quality of any runtime analysis is dependent on a number of factors. These factors include timeliness and amount of resources consumed to perform the particular analysis, among other various factors. The usefulness of the runtime analysis diminishes, in most cases, if the analysis takes too long to execute. For example, the result of a system load analyzer may not be sufficiently reliable if it takes a long time to analyze performance data. This is because the system behavior may have changed during the period of analysis, therefore when the results of the analysis become available, they no longer apply to the current running situation. Similarly, for interactive system management tasks, administrators generally expect to see the results immediately. Interactive system management tasks become a challenge for enterprise data centers or high-performance computing (HPC) systems, which have large number of possible configurations and may generate gigabytes of historical data every minute.